1. Field of the Invention
The present invention relates to a communication network system which can perform data transmissions so as to be adaptable to changes in the state of a transmission path, as well as a transmission/reception apparatus, a method, and an integrated circuit for use therein. More particularly, the present invention relates to a communication network system which employs a power line as a transmission path, as well as a transmission/reception apparatus, a method, and an integrated circuit for use therein.
2. Description of the Background Art
As communication network systems which realize data transmissions so as to be adaptable to the changing states of a transmission path by monitoring the state of the transmission path, wireless LAN systems and power line communication systems have been put to practical use.
As a wireless LAN system, IEEE 802. 11b using 2.4 GHz, and IEEE 802.11a using 5 GHz are standardized, and are widely prevalent. The above-described wireless LAN systems employ a fall down algorithm by which an appropriate modulation method is selected from among different types of modulation methods in accordance with transmission conditions. The fall down algorithm reduces a communication speed in accordance with transmission conditions. IEEE 802. 11a provides 54 Mbps transmission speed by using 64 QAM, but its communication range and noise immunity are substantially inferior to a modulation method such as 16 QAM. Thus, the wireless LAN system changes a modulation method according to transmission conditions, thereby continuing communication.
On the other hand, HomePlug1.0, which is a standard of a communication system by which 14 Mbps communication is realized by using a home power line, is developed by HomePlug Powerline Alliance, and is in practical use (see Sobia Baiget al., “A Discrete Multitone Transceiver at the Heart of the PHY Layer of an In-Home Power Line Communication Local Area Network”, IEEE Communication Magazine, April 2003, pp. 48-53).
FIG. 13 is a block diagram showing a structure of a transmission/reception apparatus 90 defined by HomePlug1.0. In FIG. 13, the transmission/reception apparatus 90 includes a transmitting-end communication control section 91, a plurality of QAM encoder sections 92, an IFFT section 93, an AFE (Analog Front End) 94, an FFT section 95, a plurality of QAM decoder sections 96, a receiving-end communication control section 97, and an SNR analytical results/acknowledgement notifying section 98.
The transmitting-end communication control section 91 determines how to allocate a bit string of input data to the QAM encoder section(s) 92 based on SNR analytical results notified by the SNR analytical results/acknowledgement notifying section 98. The transmitting-end communication control section 91 allocates a bit string of input data to each QAM encoder section 92 in accordance with an allocation scheme determined based on the SNR analytical results. That is, the transmitting-end communication control section 91 performs serial-to-parallel conversion for input data in accordance with an allocation scheme determined based on the SNR analytical results. The transmitting-end communication control section 91, which is provided with a buffer for temporarily storing input data, temporarily stores input data in the buffer. Then, the transmitting-end communication control section 91 performs serial-to-parallel conversion for the temporarily stored input data, and outputs the converted data. In the case where transmitted data is not received successfully, the transmitting-end communication control section 91 retransmits the temporarily stored input data in accordance with an acknowledgement notified by the SNR analytical results/acknowledgement notifying section 98.
A transmitting-end transmission/reception apparatus and a receiving-end transmission/reception apparatus perform a process for changing a bit allocation scheme based on the SNR analytical results in a coordinated manner. Specifically, the transmitting-end transmission/reception apparatus transmits a training packet to the receiving-end transmission/reception apparatus. In response to this, the receiving-end transmission/reception apparatus analyzes an SNR (Signal to Noise Ratio) of each carrier based on the transmitted training packet. The above-described SNR of each carrier is sent back to the transmitting-end transmission/reception apparatus as SNR analytical results. Based on the transmitted SNR analytical results, the transmitting-end transmission/reception apparatus determines a bit number allocated to each carrier. Hereinafter, the above-described process is referred to as a training session.
Each QAM encoder section 92 converts a bit string input from the transmitting-end communication control section 91 to an amplitude value and a phase value by using QAM (Quadratture Amplitude Modulation).
The IFFT section 93 executes an inverse Fourier transform based on the amplitude value and the phase value input from each QAM encoder section 92, and outputs its results. Thus, an OFDM signal modulated in accordance with the input data is output. The above-described OFDM signal is transmitted to another transmission/reception apparatus via the AFE 94.
The FFT section 95 performs a Fourier transform for the OFDM signal received from another transmission/reception apparatus via the AFE 94, and outputs an amplitude value and a phase value of each carrier.
Each QAM decoder section 96 demodulates the amplitude value and the phase value, which is output from the FFT section 95, back into a bit string by using QAM, and outputs the bit string.
The receiving-end communication control section 97 converts the bit string output from each QAM decoder section 96 to a continuous bit string, and outputs the continuous bit string as output data. That is, the receiving-end communication control section 97 performs serial-to-parallel conversion, thereby outputting output data. Also, the receiving-end communication control section 97 analyzes an SNR of each carrier based on the amplitude value and the phase value output from each QAM decoder 96 during the training session. The receiving-end communication control section 97 notifies the SNR analytical results to the transmitting-end communication control section 91 via the SNR analytical results/acknowledgement notifying section 98. The receiving-end communication control section 97 checks whether or not all packets transmitted from the transmitting-end transmission/reception apparatus are received successfully based on the generated output data. The above-described checking process is referred to an acknowledgement. The receiving-end communication control section 97 notifies acknowledgement results to the transmitting-end communication control section 91 via the SNR analytical results/acknowledgement notifying section 98.
The transmission/reception apparatus 90 shown in FIG. 13, which is compliant with HomePlug1.0, divides a data string into a large number of low rate data, and allocates the divided data to a large number of sub-carriers, each of which is orthogonal to others, for transmission. The receiving-end communication control section 97 uses a channel estimation algorithm, which is executed during a training session, for measuring an SNR in accordance with a specific frame transmitted from a transmission end. The channel estimation algorithm changes a modulation speed by estimating channel conditions. By conventional HomePlug1.0 specifications, a plurality of sub-carriers are modulated in a similar manner by selecting a single modulation parameter. However, newly performed researches have revealed that further speeding-up is realized using a method called DMT (Discrete Multitone), by which a bit number to be allocated to each carrier is determined by the transmitting-end communication control section 91 in accordance with each carrier's SNR fed back thereto.
FIGS. 14A to 14C are illustrations for describing a basic concept of DMT. In FIG. 14A, sub-carriers are denoted by numerals 1 to n, a horizontal axis indicates a frequency, and a vertical axis indicates a bit number (i.e., modulation level) allocated to each carrier. FIG. 14A shows that the sub-carriers are in the same state.
FIG. 14B is an illustration showing an exemplary SNR analyzed at the receiving end. In FIG. 14B, a horizontal axis indicates a frequency, and a vertical axis indicates an SNR value.
In the case of the SNR as shown in FIG. 14B, the transmitting-end communication control section 91 allocates a greater number of bit to a sub-carrier with a frequency of higher SNR values, and does not allocate any bit to a sub-carrier with SNR values smaller than a predetermined threshold value (SNR threshold), as shown in FIG. 14C. As such, the transmitting-end communication control section 91 controls the bit allocation scheme applied to the QAM encoder sections 92 based on the SNR analytical results, thereby changing a modulation method to transmit data without transmission errors.
The SNR is decreased by the following factors, for example: load conditions depending on a status of a device connected to a power line, noise, narrow-band noise of an amateur radio and a short-wave radio, etc., and attenuation of a signal (see Jose Abad et al., “Extending the Power Line LAN Up to the Neighborhood Transformer”, IEEE Communications Magazine, April 2003, pp. 64-70). The above-described factors change in accordance with wiring conditions, and a connection status or an operation status of a device. The factors may change on a minute-by-minute, hour-by-hour, day-by-day, or year-by-year basis.
In the conventional wireless LAN system and the power line communication system, a modulation parameter is adaptively changed by the fall down algorithm, the channel estimation algorithm, or the like. As such, a transmission speed is adjusted so as to avoid errors, thereby achieving the maximum throughput under the current transmission conditions.
In the above-described systems, a training session may be performed before communication is started. During the training session, it is necessary to perform a sequence of processes such that a specific packet (a test packet) is transmitted from a transmitting end, and a feedback packet (SNR analytical results) is sent back from a receiving end. Thus, frequent training sessions increase overhead, whereby communication speed is reduced irrespective of transmission conditions. In order to avoid such reduction in communication speed, a training session may be performed at regular intervals, for example, in a cycle of five seconds. However, the channel conditions and the above-described cycle are not synchronized. As a result, if the channel conditions change during a cycle, communication is interrupted until a next cycle is started. In the case where a training session is performed in a cycle of five seconds, for example, communication may be interrupted as much as five seconds in the worst case. Thus, even when a training session is being performed at regular intervals, a cycle thereof may be changed to an irregular cycle in the case where communication conditions are degraded due to a change in channel conditions.
In either one of the above cases, a transmitting end sends a specific packet only once during a training session, and a receiving end returns a feedback packet only once, which gives rise to the following problems.
In a power line communication system, noise and impedance fluctuates in synchronization with a supply power cycle or half the supply power cycle. For example, in the case where the supply power cycle is 50 Hz, the noise and impedance will have a fluctuation cycle of 20 msec or 10 msec. In the case where the supply power cycle is 60 Hz, the noise and impedance will have a fluctuation cycle 16.7 msec or 8.3 msec.
Noise which is in synchronization with the supply power cycle or half the supply power cycle of a home appliance, and impedance fluctuations during a supply power cycle of a half- or full-wave rectifier circuit in a home appliance are causes of noise/impedance fluctuations. Noise/impedance fluctuations occur locally, in the neighborhood of the responsible home appliance or the like. In particular, it has been found that large fluctuations in noise and impedance may be ascribable to a recharger for a cellular phone, an electric carpet heater, or the like. Since these devices are in wide use at general households, it is necessary to devise countermeasures against the fluctuations in transmission path characteristics caused by such devices, in order to be able to transmit AV (Audio Visual) signals, which require low-delay and low-jitter characteristics.
FIG. 15 is a graph showing temporal changes in the phase of a signal which is transmitted over a power line to which an electric carpet heater is connected. As shown in FIG. 15, in accordance with the impedance fluctuations of the power line due to a half-wave rectifier circuit provided in the electric carpet heater, the phase of the signal being transmitted over the power line is fluctuating with a cycle of about 8 msec.
FIG. 16 is a graph showing measurement results of SNR of each carrier, obtained through channel estimation in the presence of power line impedance fluctuations as shown in FIG. 15. FIG. 16 show fluctuations in the measurement results of SNR of each carrier which are obtained through channel estimation in the presence of power line impedance fluctuations. A frame in which the SNR shows a maximum value is a frame which is transmitted at a point in time where the phase of the signal being transmitted over the power line is not fluctuating. A frame in which the SNR shows a minimum value is a frame which is transmitted at a point in time where the phase of the signal being transmitted over the power line is fluctuating. There is a maximum difference of 20 dB in the SNR of sub-carrier numbers 120 to 200 (corresponding to frequencies of 10 to 15 MHz) between the frame transmitted when the phase of the signal being transmitted over the power line is fluctuating and the frame transmitted when the phase of the signal being transmitted over the power line is not fluctuating. Thus, there can be as much as 20 dB of SNR fluctuations, depending on whether or not the point in time where a channel estimation is performed coincides with the point in time where the phase characteristics undergo a great fluctuation.
Next, the relationship between SNR evaluation results and communication rates will be discussed. FIG. 17 is a table which shows off-line simulation results of communication rates in the physical layer (hereinafter referred to as “PHY rates”), with respect to a maximum SNR and a minimum SNR, in the case where a channel to be used and a modulation level are selected based on a channel estimation.
As shown in FIG. 17, there is a difference of 30 Mbps between the PHY rate selected for the case where the SNR is maximum and the PHY rate selected for the case where the SNR is minimum. Therefore, in the conventional technique where a channel estimation is performed only once during a training session, if the transmission timing for an estimation request packet coincides with a phase-fluctuating moment, a PHY rate which is based on the minimum SNR will be selected. In this case, the communication efficiency is deteriorated because, during a period where the phase is not fluctuating, a transmission rate which is 30 Mbps slower than an actually available PHY rate is being used.